Research On Digitally Reconstructed Radiograph Rendering Method And Its Application

Radiotherapy has been become an important method of malignant tumor treatment. Treatment Planning System (TPS) is used for pre-operative planning of radiotherapy and has a significant influence on the accuracy and effectiveness of radiotherapy. It is one of core technologies of radiotherapy. The Digitally Reconstructed Radiograph (DRR) rendering technology generates similar X-ray image-like image from digital CT data by simulating the attenuation and exposure process of X-ray. DRR has been widely used in TPS for portal field verification, patient position system. Based on self-developed"Advanced / Accurate Radiation Therapy System"(ARTS), the dissertation focuses on DRR rendering method and its application in ARTS, main contents include:(1) ITK Based DRR Rendering Algorithm: ITK is a powerful and open-source development package for medical image processing, which is used in ARTS. This dissertation presents an ITK based DRR algorithm, which has been applied to ARTS effectively.(2) CUDA Based DRR Rendering Algorithm: On viewpoint of visualization, DRR rendering algorithm is a kind of direct volume rendering algorithm. Because Direct volume rendering is computation-intensive algorithm the dissertation presents a CUDA based DRR rendering algorithm for acceleration.(3) Depth Peeling Based DRR Mixed Rendering Algorithm: The DRR rendering algorithm is a direct volume rendering algorithm. However, when the volume data (CT slices) is rendered in TPS, the meshes of organ, target volume and Multi-Leaf Collimator should be displayed simultaneously. For this reason, this dissertation presents a Depth Peeling based DRR mixed rendering algorithm. The theory of the algorithm and the programmable GPU based implementation which achieves satisfied results are both demonstrated in this dissertation.